50th Anniversary of the Mother of All Demos (2018)
amasad.me
amasad.me
He must’ve been terribly disappointed with how things evolved. In the 1960s, he envisioned computers to keep track of all human knowledge for us, to help us find contradictions, fill knowledge gaps, and ask better questions. And what did we get instead? Clickbait and spam bots. Did he see it that way?
The fact computers have been so isolated from one another (even with internet apps) is a complete tragedy. Even sending large files is a pain in the butt to this day.
The demo had multiple people working on the same computer remotely, with their own cursors and ability to do things on the system at the same time.
Best we have now is web-browsers syncing data silos, or a one way "let this person control my mouse + keyboard".
It feels like building a bridge by everyone making parts in isolation then expensively shipping it to the work site that only has one crane to assemble them.
Usually it's way easier to have multiple people work on something. Like carrying something heavy, or having a line of people walk the bridge with the right tools/verification to install the rivets right, or even the opportunity for someone to speak up present unique ideas to the group.
Even just a few basic OS primitives early on could have jump started so much collaboration tools and techniques that mirror how we all work together everywhere else in real life.
As an aside, the whole "meta-verse" craze might pull these ideas through to the mainstream again.
I see lots of demos of avatars standing at a table working on stuff together, and I've been on VR chat before drawing 3d images with crayons at the same time with other people.
It's incredibly fun to try and draw an elephant together and work on different parts at the same time to hilarious results.
> Right now we’re in the era of personal computers. Collaboration, social use of tools, togetherness: all of these are hacks on top of something that, at its core, was designed for the individual first.
> But there’s a particular photograph of group computing from the 1950s, from before the PC was invented...
I wonder how much if at all Engelbart was aware of SAGE. Webb poonts out Engelbart got started working on computers out of SAGE.
> As an aside, the whole "meta-verse" craze might pull these ideas through to the mainstream again.
Heck yeah, right on. Even if the immersive modes fail, I really hope some embodiment survives. Constructing space & environments in which being & event can occur, manifesting processes... I have a lot of hope we can start to better ground & make real computing.
[1] https://interconnected.org/home/2021/12/21/sage https://news.ycombinator.com/item?id=31665364 (12 points, 10 hours ago, 2 comments) https://news.ycombinator.com/item?id=31535139 (14 points, 12 days ago, 1 comment)
That's amazing lol
Apple's multi-user support in it's "iWork" productivity apps actually comes very close to this, and it's a real shame that it's not as well known as it is. In fact Apple just demo'd improvements to it a couple of days ago at WWDC.
https://support.apple.com/en-us/HT206181
It works almost exactly like you're describing and frankly I think it's amazing and am surprised more people don't rave about it.
One of the few reasons why I have Windows installed on my machine was so that my ex-girlfriend could use my computer when she needed to. Somewhat amusingly, I had to teach her how to access the arrow keys under a function layer of my 60% keyboard to select the correct bootloader entry in order to do that. That's two new concepts foreign to most average users just to do what they might consider "turn on the computer".
Having said that, some implementations of collaboration don't have this problem. In most modern networked applications that are designed to be multi-user, each client runs its own instance of the UI, and the clients synchronize the underlying data. This scheme can accommodate different user needs much better than, say, using something like VNC to have a bunch of people watch a shared frame buffer and take turns sending input to a shared UI.
(For what it's worth, it's also available on any platform via the web app versions of iWork at icloud.com. I'm not saying this is necessarily practical, but, it's there.)
It really all hinges on how you want to define "computer". There are some prerequisites like you need a network connection and display and input device(s). You also need some client software. In this case, the client consisting of a particular software stack doesn't seem that unreasonable.
If I can't play with you using the computer in my laser printer it doesn't mean the vision of collaborative computing still isn't here.
MS Office supports this, as does Google Docs, and apparently, so does iWork!
I actually really agree with this hypothesis. People think the whole head mounted display thing is going to revolutionize interaction (and I do think eye contact will be a killer app)…
But I also think one of the biggest affects VR will have is to reset peoples’ expectations about 2D interfaces. I think once we get used to the idea that “data exists in space” it’ll trickle down into the 2D world and—50 years after it began was invented—we might see an alternative to WIMP.
I vaguely remember some multi-user system for Windows as well, but IIRC it was more of an R&D thing.
We don't have to really send anything anymore, I can just place a file in a shared location. I can even share an incredibly high resolution video on google by just making a link.
>Best we have now is web-browsers syncing data silos, or a one way "let this person control my mouse + keyboard".
I am now able to edit Excel documents and see people in the same file making updates live. I am not saying this is peak collaboration but I find it pretty impressive.
The meta-verse may have a chance (I'm pulling for it in some shape), but let's not act like we aren't making progress :)
But I am left wondering what could have been, what might have been. Nobody would fund him. Apple probably should have just given him a lab with a couple of employees and a modest budget. Who knows what he might have invented? He had that rare ability to see into the future. The world lost a good thirty years of his genius and not being a Valley insider I am left wondering why?
He was full of interesting ideas, not someone who just had one big insight long ago, and was great to talk to about your own work.
His memorial service was full of amazing people whom he’d met and worked with in the Valley over the decades. Mind boggling how that works — quite different from a fully remote world. We talk about WFH vs office, but less about the value of running into Doug engelbart or don knuth while walking down the street.
I was at that service and several people spoke about how his ideas did not really get the backing they deserved, Ted Nelson not the least of them.
This makes more sense when you realize that what Doug really wanted was to transform the way people worked far more than he wanted to invent cool new tech, but while people were happy to have him as a member of staff he needed to be given freedom to play with organizations and no one was ready to try anything as radical as he had in mind.
I remember there was a period where Google was trying to make him a permanent fixture and he said in a private conversation something to the effect of it feeling like being surrounded by a bunch of young people who just wanted to admire him like a statue, not engage with his ideas.
Yeah, people were at the center of his thinking, not machines. Again, crazy farsight by today’s standards.
> feeling like being surrounded by a bunch of young people who just wanted to admire him like a statue
This is hard: classic imposter syndrome makes ppl feel afraid to ask a dumb question.
> not engage with his ideas.
A separate problem; some of the drivers of this I touched on in another comment; would be interesting to read what you think.
I remember feeling that Ted was talking about himself, as he chronically does. Ted’s good, radical, and farsighted ideas didn’t get the attention they deserved either, but I feel a lot of that was self inflicted. And the amount they deserved might have been less than he thought.
Come to think of it, Ted should recast trans Lu soon as a web3 idea. His charge-for-everything approach would fit right in.
I'm quite confident he was talking about both Doug and himself. Ted definitely saw Doug as a kindred spirit with a great vision for how the world could be different who was ignored.
While we've evolved to value in-person interactions higher, let's not forget the fact that even non-sonic non-visual near-real-time conversation with actual creator would be simply impossible for someone sitting in a third-world country, like me :)
This is perhaps the defining sadness that most saps at me. This is a world of such potential that does such expensive vast things. I really wish there were more room & space for people to go off & explore, try testing new novel spaces.
Dont we have some surplus, to feed the promising with, to give them a couple allies to go boldly sally forth with?
The corporation as a model has such deficiencies, has such need for alignment & concensus & purpose. Humanity is just so much greater when there's some permissionless innovation, some real testing & pushing things just to find out, to try & see, to go further. Our models of labor & vue seem so unable to make small but long bets on so many bright stars.
> "Engelbart slipped into relative obscurity by the mid-1970s. As early as 1970, several of his researchers became alienated from him and left his organization for Xerox PARC, in part due to frustration, and in part due to differing views of the future of computing. Engelbart saw the future in collaborative, networked, timeshare (client-server) computers, which younger programmers rejected in favor of the personal computer. The conflict was both technical and ideological: the younger programmers came from an era where centralized power was highly suspect, and personal computing was just barely on the horizon."
Now both concepts have become more-or-less fully developed with a wide variety of small personal computers from RPis to desktop powerhouses, as well as rent-space-in-the-cloud client/server approaches. However, you do need to have a lot of faith in the overall power structure not going down some authoritarian rat hole to rely solely on the latter.
His ideas are certainly interesting and still relevant today:
> "Consider a group's Collective IQ to represent its capability for dealing with complex, urgent problems – to perceive and understand them adequately, to engage the stakeholders, to unearth the best candidate solutions, to assess resources and operational capabilities and select appropriate solutions and commitments, to be effective in organizing and executing the selected approach, to monitor the progress, to be able to adjust rapidly and appropriately to unforeseen complications, and so on."
https://futureofcoding.org/episodes/056
That paper (from '62) forms quite the contrast with MoAD (from '68). While the latter is quite specific, showing a lot of "features" that people tend to point to when citing Engelbart's influence, the former is arguably much more visionary. Doug had quite a lot more in mind than just pointing devices and video calling.
> [... 1:17:32] Joe is writing down all of the things that you’re saying describing the system, and then he says that “it represents the clumsy phrasing and illogical progression of thought so typical of extemporaneous speech” [https://www.dougengelbart.org/pubs/augment-3906.html#3b5a] which A: is what Jimmy and I are doing right now, B: is what Doug Engelbart’s writing reads like anyways. His writing reads like clumsy and illogical progressions of extemporaneous speech, except that it’s like, worked to within an inch of its life. It is so overwrought, and yet at the same time clumsy and illogical, so that... you know... grrrr!
> [... 1:18:14] Doug is a bad writer, I don’t like Doug’s writing, I’ll say that, I have not enjoyed reading his writing. I enjoyed his fan fiction better than I enjoyed his technical manuals, it’s just not a pleasure to read. And it made me wonder, [...] did he feel the disconnect in quality between, like, Bush’s writing and his own writing? Did he feel like the quality of his writing needed to be formed into this very unappealing mold because of the context he was working in?
> [... 1:20:42] I do kind of agree with you, his writing is very awkward, right? It is hard to kind of get at the ideas that are in here, and you know, maybe there is something to it where he is... this is clearly, some of this is an intentional stylistic choice, right? [...] he wanted to write in this way for some particular reason or other, I think.
Thank you for this call out. This idea, of whether computers are here to mechanize processes that people fit into, or whether computers are tools to augment & enhance our individual agency, is a huge question.
Ursala Franklin has some good material on Technological Society. There's a couple different points of view she proposes, but I really enjoy the distinction she makes between Holistic technologies, where the capabilities of the user are expanded & opened, the user gains control, versus Prescriptive technologies, which relate more to macro-level control, to enforcing set-down processes[1].
[1] https://en.wikipedia.org/wiki/Ursula_Franklin#Holistic_and_p...
I was there because I read every one of his group's papers, in the mid 80's, and wrote a version of Augment in MS-DOS and Turbo Pascal. I think I showed him a demo.
He invited me to attend informal meetings at SRI. I left after a few weeks because I couldn't figure out how it would result in a product. Had already went into deep personal debt, once before (about $50K in 1990), trying to market something similar.
But, then again, not sure how you put a coin slot on the Arpanet.
Am still working with core concepts: using a namespace hierarchy (akin to a WBS) and a directed graph (akin to activities). Except now, the activities are code.
He wanted to see how much he could optimize their productivity over time with careful measurement and plain old trial and error.
It's never stated why funding dried up, but here's my suspicion: Nobody was willing to risk upsetting the traditional hierarchy with something that might obsolete management, for example. Nobody other than an angel investor, or group of old retired technologists can pull something like that off. It has to be someone with time, or f*ck you money. I have time.
https://en.wikipedia.org/wiki/The_Mother_of_All_Demos
I met Doug in the late 1970s when we worked in the same Tymshare office at 20705 Valley Green Drive in Cupertino. He had a little cubicle like the rest of us, in the far corner of the building behind the pine tree in the middle of this Street View image:
https://goo.gl/maps/JEZAqS6vqVAAfjP6A
We only talked a few times, but I could always see the sadness in his eyes from the shabby treatment Tymshare gave him. No real team, and not much encouragement to continue his mission.
Tymshare was quite a land of lost opportunities. Earlier in the 1970s they had a "programmable terminal" in a side office. I think it may have been a Datapoint 2200?
I found the programming manual and saw that I could write code to respond to the cursor keys, put characters on the screen where I wanted them, and then run Conway's Game of Life!
In my years of working there coding on Teletype machines, I'd never experienced anything like this.
Eventually my manager saw me working on my GoL program, and kindly but firmly said, "Mike, this looks like fun and all, but no one is ever going to want to do something like this. We have important business problems to solve. Back to work now!"
At one point they did consider doing something to take advantage of these newfangled "microcomputers". But they didn't pursue it very far, because they figured it might be a threat to their established timesharing business.
And they were right!
How did something so monumental get built and demonstrated, but then not get followed up on for more than 30 years?
Were there humongous caveats involved in the demo that made it impossible to leave the lab even for major corporations?
And the labs in those days were pushing forward and “inventing the future”. We had network-native systems: not just computers with network adaptors, but cloud native software with mail, filesystem etc transparently available over the net. In fact where did the “cloud” concept come from? A TCP design document in the 70s.
When you’d look outside the lab it felt like the past. DOS was the state of the art. And you’d visit your friends at different institutions and see different approaches to address the same problems.
These days I never get that “this is the exciting future” feeling. The old nexus corporate labs (PARC, IBM, Bell Labs (RIP) and others) aren’t dominant as they used to be, but when I visit their successors (e.g. Google) the research doesn’t have that same jaw-drop impact. Not because I’m jaded, but because nothing appears to have come back through a time machine any more. Perhaps because computing went mainstream and then the constraints appeared?
* and are: I think people still haven’t grasped how transformational they will be once we finally get computers to work and be useful. We’re still just puttering around at the base of the mountain.
> The caveats were that 1 - you needed specialized research hardware that basically nobody had and 2 - people didn’t understand how revolutionary computers were.* Even in the late 80s the general consensus was that personal computers would be just used by specialists because “no executive can type, or would want to”
I agree with this completely. The teleconferencing, for example, required a big string of trucks with equipment to transmit the signal IIRC. An incredible tech demo but not even close to something that could be productized.
> Not because I’m jaded, but because nothing appears to have come back through a time machine any more. Perhaps because computing went mainstream and then the constraints appeared?
I think another way to frame this is that early computers offered a clear paradigm shift, the shift happened, and now everything on the near horizon is within that paradigm. This true even for some of the more exciting projects like Bret Victor's dynamicland (I'm not sure it will actually go anywhere, but I'm happy to see someone carrying the torch of the multimedia dreams that died in the 90s).
Personally, I think biotech has the best potential for the next "it's the future!" style revolution.
SpaceX is the company I look at as clawing the future back to us piece by piece.
As for caveats, Sutherland was using a $1.3m computer (2022 money) of which only 50 existed, and two TV microwave links ($$$) which are impossible to scale.
Still, we can dream.
Same as with moon shots. Abandoned.
https://www.dougengelbart.org/pubs/augment-3906.html
It's worth noting that despite the historical magnanimity of the 68' demo that the real, "meat and potatoes" of Doug's work is still the operating philosophy he developed that allowed SRI to dream up all this groovy shit and make it work. It's also worth remembering Bill English who was the, "director" of the demo who sometimes doesn't get enough credit.
Yes, as Doug Engelbart said when talking about the demo: “but anyway it worked, and the main reason it worked is because Bill English is a genius”.
>The 90-minute presentation demonstrated for the first time many of the fundamental elements of modern personal computing: windows, hypertext, graphics, efficient navigation and command input, video conferencing, the computer mouse, word processing, dynamic file linking, revision control, and a collaborative real-time editor
If those things were possible in 1968, why did most of us started to use them just in the '90?
Is normal to have that big of a gap between a technological demonstration and the technology becoming available?
So it's not like people forgot about it for a couple of decades, there's a clear line from that demo to where we are now.
Had the PC not won it's possible that something like multi-user systems would have become the main thing (which we now basically have, but we call it "the cloud").
"Personal" is the pair of shoes I wear every day, even though they're years past their use-by date, are ugly, smell funny and the soles are worn. For you it might be the jumper knitted for you by your Nan. It's the objects we use that, over the years, mould themselves to our personal idiosyncrasies, habits, preferences and style. Our computers scarcely have that (particularly in the Apple and MS worlds of Personal Computing!) and hardly live long enough to acquire personalisation/personality. Not to mention, too, it's quite hard to do. Example: my (Linux, KDE) box is pretty happy for me to define custom hot-keys for all sorts of things, from firing up frequently-used apps (I scarcely ever use the desktop menu these days) to typing my email address. But could my (tech-clueless) brother do that? Extremely unlikely. And it takes several non-intuitive steps to make a new hotkey...
"Computer": How programmable is my phone? (I've done Android dev work. I don't ever wish to do it again.) How scriptable? How malleable in casual ways? Not at all. It's Google's device much more than it's mine. It's a sort of locked-down surveillance television, not a general-purpose computer programmable-by-me (except in a highly specialist and narrow sense.)
Much of the blame I put at the doorstep of Visual Basic which reduced the scope and utility of direct-manipulation interfaces into a terribly narrow, constrained version of "paper forms on an electronic screen". It put the vision back by at least a decade. Maybe several.
Finally, though, I have hope we might emerge from that. There's a bunch of us working on it.
The de-facto deification is pervasive.